Structural and Functional Comparison of HDL From Homologous Human Plasma and Follicular Fluid
A Model for Extravascular Fluid
Abstract In the preovulatory period, follicular fluid contains only HDL. Biochemical characterization of such lipoproteins showed that follicular fluid HDLs were cholesterol-poor particles compared with serum HDLs, whereas the amount of phospholipids, expressed as percent weight, was significantly higher in follicular fluid HDLs (28.5%) than in serum HDLs (25.0%, P<.05). The amount of apolipoprotein (apo) A-IV per apo A-I was significantly higher in follicular fluid than in serum (0.77 versus 0.58 mg/g apo A-I, P<.02). To explore the role of HDLs as cholesterol acceptors in physiological media, we compared the ability of either whole human follicular fluids or homologous sera to promote cellular cholesterol efflux using Fu5AH rat hepatoma cells. At equivalent concentrations of HDL cholesterol in follicular fluid and in serum, t1/2 values for cholesterol efflux were in the same range. In addition, estimated maximal efflux values were not significantly different in follicular fluid and serum (45.9% and 49.6%, respectively), as were Km values (0.064 and 0.071 mmol/L HDL cholesterol, respectively). In addition, isolated HDLs displayed the same capacity to promote cellular cholesterol efflux in both media. Thus, the kinetics and dose-response data between these two physiological media showed that HDLs play the major role in cellular cholesterol efflux. The rate of cholesterol esterification, as measured in the presence of cells, was significantly higher in follicular fluid than in serum at constant HDL cholesterol concentrations, whereas the rate of esterified cholesterol transfer toward added LDL was lower. In contrast, in a cell-free system, lecithin:cholesterol acyltransferase activity represented only 26% of that in serum HDL, whereas cholesterol ester transfer protein activities were comparable. In summary, in this particular model, we confirmed the essential role of HDLs as physiological acceptors in the removal of cellular cholesterol.
- Received July 25, 1996.
- Accepted November 8, 1996.
HDLs are involved in reverse cholesterol transport from peripheral cells to the liver,1 and this process primarily implicates HDL particles, which are considered to be the physiological acceptors for cellular cholesterol efflux, the first step of reverse cholesterol transport. A number of cellular models have been developed to elucidate the role and relative importance of HDL subspecies in cellular cholesterol efflux. Most of the studies were performed using HDL particles isolated from serum or reconstituted particles containing apolipoproteins and phospholipids2 3 4 5 and have provided information on the biochemical and physical factors that influence the efficiency of acceptor particles to promote efflux.6 In addition, studies using whole plasma or serum7 8 9 have established that different acceptors participate in the modulation of cell cholesterol efflux, but in such complex media that contain a mixture of lipoproteins, it is quite difficult to distinguish the role of HDL particles from those of other lipoproteins. Interestingly, in contrast to the plasma and the interstitial fluid that contain VLDL,10 11 the human follicular fluid obtained from the preovulatory follicle contains a single class of lipoproteins, HDL.12 13 14 Human follicular fluid, which surrounds the preovulatory oocyte in the ovarian follicle, is in direct contact with granulosa cells but separated from the thecal blood capillaries by the basement membrane (for a review, see Schreiber and Weinstein15 ). Hence, follicular fluid constitutes a model of extravascular fluid. As previously suggested,14 16 17 18 this extravascular fluid might originate, at least in part, from blood plasma transported through the follicle basal membrane, since major follicular fluid proteins are plasma proteins. During the preovulatory period, the tissues and the basement membrane become permeable to serum proteins up to 3×105 Da in size.16 This can be explained by the porosity of the follicle membrane, which is sensitive to hormonal status.
In the present study, we characterized the apolipoprotein and lipid compositions of HDL from follicular fluid and homologous human plasma. Furthermore, we tested the ability of this extravascular fluid, compared with homologous sera, to promote cholesterol efflux from Fu5AH rat hepatoma cells to better understand the role of HDL in this first step of reverse cholesterol transport.
Fatty acid–free bovine and human serum albumin, aprotinin, leupeptin, and egg yolk lysophosphatidylcholines were obtained from Sigma Chemical Company. 125I-labeled goat anti-mouse F(ab′)2, 125I-labeled donkey anti-rabbit F(ab′)2, and [3H]cholesterol were purchased from Amersham. Monoclonal antibodies against human apo A-I were kindly provided by Dr Ross Milne and Dr Yves Marcel (Ottawa, Canada). Rabbit polyclonal antibodies against human apo A-IV were obtained from Dr Para (Institut Pasteur). Commercial kits of cholesterol oxidase/esterase and apo B/E precipitation reagents were obtained from Boehringer and BioMerieux. Calibration standards for exclusion chromatography and polyacrylamide gel electrophoresis were obtained from Bio-Rad and Pharmacia, respectively.
Follicular Fluids, Sera, and Plasma Collections
Preovulatory follicular fluids and sera were obtained from patients enrolled in an in vitro fertilization program, as previously described.19 Hyperstimulation was achieved with a two-step procedure. The first step consisted of pituitary blockade using luteinizing hormone–regulatory hormone analogues, which abolish the endogenous secretion of follicle-stimulating hormone and luteinizing hormone. The second step consisted of follicular stimulation, which was achieved by injection of human menopausal gonadotropins. Preovulatory follicles were punctured during laparoscopy. Only the cleanest follicular fluids, ie, those with no evidence of blood cell contamination, were retained.
Isolation of Follicular Fluid, Plasma, and Serum HDL
HDL particles were obtained from follicular fluid, plasma, and serum after the addition of an anti-protease mixture containing EDTA (0.2 mmol/L), sodium azide (0.01%, wt/vol), PMSF (0.1 mmol/L), iodoacetamide (1 mmol/L), 1,10-phenanthroline (1 mmol/L), leupeptin (0.1 mmol/L), and pepstatin A (1 μmol/L). VLDL and LDL were precipitated by adding phosphotungstic acid/magnesium to the samples (0.5 mL). Despite the absence of apo B/E–containing lipoproteins in follicular fluid, the precipitant was added to create the same condition as in plasma and serum. After centrifugation (550×g for 10 minutes), HDL particles were recovered in the supernatant. The HDL particles were then isolated from supernatant by ultracentrifugation at a density of 1.21 g/mL and 120 000×g for 40 hours at 7°C. The lipid and apolipoprotein concentrations of the HDL particles were measured as follows.
Extraction and Analysis of HDL Neutral Lipids by Gas-Liquid Chromatography20
Lipids of each sample were extracted according to the method of Bligh and Dyer21 after acidification of the aqueous phase with formic acid (12 μL/mL). Before extraction, four internal standards were added on the basis of 20 nmol of free cholesterol extracted: 3 μg of stigmasterol, 6 μg of 1.3 dimyristoyl-sn-glycerol (DG14:0), 6 μg of heptadecanoyl cholesterol (CE17:0), and 1 μg of triheptadecanoyl glycerol (TG17:0), all from Sigma. After evaporation of the chloroform phase to dryness, extracts were dissolved in 100 μL of ethyl acetate. Lipids were analyzed by gas-liquid chromatography (Intersmat, model 120 DFL) using an Ultra 1 Hewlett-Packard fused silica capillary column (5 m×0.31 mm inner diameter) coated with cross-linked methyl silicone. Oven temperature was programmed from 205 to 345°C at a rate of 6°C/min, and the carrier gas was hydrogen (0.5 bar). The response factors for the different lipid classes were determined using a mixture of internal standards. The variation coefficient of intra- and interassays represented less than 6%.
Quantitation of Phosphatidylcholine Free Fatty Acids
Lipids of HDL, obtained after precipitation of apo B/E–containing lipoproteins, were extracted according to the Bligh and Dyer21 technique. Before extraction, an internal standard of 15 μg of diheptadecanoyl phosphatidylcholine (PC C17:0) was added to the samples. Phospholipids were separated by thin-layer chromatography on Silicagel G60 using a mixture of chloroform, methanol, water, and acetic acid (75:45:6:12). The phosphatidylcholine spot was scraped off. Transesterification of phosphatidylcholine was accomplished by adding a mixture of acetyl chloride and methanol (1/20) to the silica, followed by 1 hour of incubation at 55°C in a sand bath. Methylated free fatty acids were then extracted from the silica by addition of water and spirit ether (1:1), followed by centrifugation at 1100 rpm for 5 minutes. After evaporation to dryness, samples were dissolved in 150 μL of ethyl acetate and analyzed by gas-liquid chromatography (Carlo Erba) using a polar fused-silica capillary column coated with poly(ethyleneglycol) Stabilwax (30 m×0.32 mm inner diameter). Oven temperature was programmed from 160 to 220°C at a rate of 1.8°C/min, and the carrier gas was hydrogen (0.5 bar). Molecular species were resolved on the basis of their carbon number and degree of unsaturation. The response factors for the free fatty acid classes were determined using the internal standard.
Rates of Cholesterol Esterification and Transfer in a Cell-Free System
Cholesterol esterification and esterified cholesterol transfer activities in follicular fluid and homologous serum (n=4) were assayed in a cell-free system as follows. VLDL and LDL were precipitated by adding phosphotungstic acid/magnesium to the samples (0.5 mL). Despite the absence of apo B/E–containing lipoproteins in follicular fluid, the precipitant was added to create the same condition as in plasma. After centrifugation and dialysis to remove any phosphotungstic acid/Mg2+ reagent, the HDL-containing supernatants were labeled with [3H]cholesterol as follows. [3H]Cholesterol (106 dpm) was spotted onto Whatman 3M paper (previously washed with HCl 0.1 N/ethanol). Supernatants containing HDL (1.2 mL) were incubated with [3H]-labeled paper for 16 hours at 4°C. At the end of labeling, an aliquot was counted, and an average of 27% of total radioactivity was recovered in the follicular fluid and plasma HDL fractions (range, 22% to 32%, respectively). To measure the esterification rates, samples were further incubated at 37°C for 2 and 4 hours. Aliquots (0.3 mL) from each incubation time were extracted according to the method of Bligh and Dyer.21 Esterified and unesterified cholesterol were separated by thin-layer chromatography. The esterification rate was calculated and expressed as nanomoles of esterified cholesterol formed per hour per milligram of apo A-I.
After the esterification period, an assay to measure esterified cholesterol transfer in samples was performed after addition of LDL (LDL/HDL cholesterol molar ratio, 4:1) and 0.4 mmol/L 5,5′-dithiobis-(2-nitrobenzoic)acid (as an inhibitor of LCAT activity). Samples were incubated for 16 hours at 37°C, and LDL was precipitated as described above. HDL and LDL lipids were then extracted. Esterified cholesterol and unesterified cholesterol of HDL and LDL were separated by thin-layer chromatography. The esterified cholesterol transfer activity was expressed as nanomoles of esterified cholesterol transferred from HDL to LDL per hour per milligram of apo A-I.
Kinetics and Dose Response of Cellular Cholesterol Efflux
Fu5AH rat hepatoma cells were grown and labeled as previously described by de la Llera Moya et al.8 We compared the kinetics of cholesterol efflux from the cells with the kinetics of cells from a pool of follicular fluids (n=8) and a pool of homologous sera (n=8) at the following concentrations: 1%, 2.5%, 5%, 10%, 25%, and 50% (vol/vol). We did not exceed 50% for each sample to prevent cellular toxicity. For each concentration and sample, we measured the fractional cholesterol efflux at the following incubation times: 10, 20, 30, 40, 60, 120, 180, and 240 minutes. All experiments were done in triplicate. After incubation, follicular fluid or serum containing medium was removed and kept at −70°C before analysis. To standardize the cellular response obtained with different batches of cells and labeling media, a standard serum pool was always included as a test serum. At the end of the efflux experiment and after cell monolayers were washed with phosphate buffer, cellular lipids were extracted by incubating the cells with isopropanol (overnight at room temperature). Fractional efflux was calculated as the ratio between the label released to the medium and the total radioactivity recovered in media plus cells.
Plots of ln(1−FE) as a function of incubation time were established for each sample concentration, where 1−FE represented the remaining [3H]cholesterol in cells, which allowed us to determine the apparent t1/2 (half-time for removal of cholesterol) as the time for 50% cholesterol efflux. Then, the t1/2 values were plotted versus the total HDL cholesterol concentration in the follicular fluid and in the pool of homologous sera.
Dose-response data were obtained from an incubation time of 4 hours with follicular fluid and serum and were linearized according to the Lineweaver-Burk plot to determine the estimated maximal efflux (Vmax efflux) and the Km values for each sample.
Cellular Cholesterol Efflux, Esterification Rate, and Cholesterol Transfer Assays
The efflux capacities of whole human follicular fluids, whole homologous sera, and HDL isolated from homologous serum were assayed by incubating individual samples, at a concentration of 5% (vol/vol) or, depending on the experiment, at a constant concentration of HDL total cholesterol, with radiolabeled cells for 4 hours at 37°C. Serum depleted of VLDL and LDL, ie, HDL supernatant, was obtained under the precipitation procedure described in “Methods” (see “Rates of Cholesterol Esterification and Transfer in a Cell-Free System”). We also performed experiments to directly test the effect of isolated HDL from a pool of follicular fluid and homologous serum (n=6), used at a constant total cholesterol concentration (0.075 mmol/L), on the cholesterol efflux from Fu5AH rat hepatoma cells during incubation times varying from 30 to 480 minutes. These HDL particles were isolated as described in “Methods” (see “Isolation of Follicular Fluid, Plasma, and Serum HDL”). In some studies, we compared the ability of follicular fluids (used at 5%) to promote cellular cholesterol efflux to that of follicular fluids supplemented with LDL, added at the same concentration found in plasma (LDL/HDL cholesterol molar ratio, 4:1). The fractional efflux from cells to the samples was calculated as described above.
The esterification rate of the efflux medium, used as an estimation of the functional LCAT activity, was measured by quantitating the labeled cholesterol esterified in each medium sample during the efflux period.8 After incubation of cells with different samples, an aliquot of medium was extracted according to the method of Bligh and Dyer.21 Esterified cholesterol and unesterified cholesterol were separated by thin-layer chromatography using a mixture of petroleum ether, diethyl ether, and acetic acid (98:2:1), and the radioactivity was counted. Finally, the esterification rate was determined as the ratio between esterified cholesterol and total cholesterol. The molar cholesterol esterification rate was calculated on the basis of specific radioactivity and ranged from 5000 to 10 000 dpm/nmol cholesterol.
The rate of cholesterol transfer between HDL and apo B/E–containing lipoproteins was measured and used as an estimation of the CETP activity in the efflux medium.8 Apo B/E–containing lipoproteins of the efflux medium (0.8 mL) were precipitated by adding 10% phosphotungstic acid/MgCl2 and bovine serum albumin as carrier protein (7 mg/mL). After centrifugation, an aliquot of the supernatant containing labeled HDL was counted. The lipids of precipitated apo B/E–containing lipoproteins were extracted according to the method of Bligh and Dyer,21 and esterified cholesterol and unesterified cholesterol were separated by thin-layer chromatography, as described above. The rate of cholesterol transfer was determined as the ratio between esterified cholesterol in the precipitate and total esterified cholesterol in the efflux medium.
Free and total cholesterol were measured by cholesterol esterase/cholesterol oxidase techniques. Measurement of proteins was performed according to the method of Lowry et al22 using bovine serum albumin as a standard. Lipid extraction was performed as described by Bligh and Dyer21 after acidification of the aqueous phase by formic acid (12 μL/mL aqueous phase). Phospholipids were measured according to their phosphorus content23 after lipid extraction. Different classes of lipids were separated by thin-layer chromatography on Silicagel G60. For the efflux studies, esterified cholesterol and unesterified cholesterol were separated on polysilic acid gel–impregnated glass fiber.8 The phospholipid separation was conducted as described by Skipski et al24 using a mixture of chloroform, methanol, water, and acetic acid (75:45:6:12). Results are expressed as the mean±SE. Statistical comparisons were performed using Student’s t test for paired samples. Apo A-I and B were measured by laser immunophelemetry using specific antisera against purified human apoproteins. Lipoprotein particles Lp A-I and Lp A-I:A-II were measured by immunoelectrodiffusion.
Lipoprotein, Apolipoprotein, and Lipid Composition in Human Plasma and Follicular Fluid
In female plasma, the mean total cholesterol concentration was 4.8±0.2 mmol/L; in contrast, the total cholesterol concentration in homologous follicular fluid (n=8) was only 0.8±0.1 mmol/L, which reflected the absence of LDL or VLDL (not shown). In addition, the absence of detectable apo B confirms the absence of such lipoproteins. Thus, the majority of cholesterol was recovered in the range of HDL lipoproteins, as previously described.14 The concentrations of apo A-I and A-II are significantly lower in follicular fluid HDL than in plasma HDL, whereas the amount of apo A-IV is not significantly different (Table 1⇓). However, when expressed per gram of apo A-I, the amount of apo A-IV was significantly higher in follicular fluid HDL than in serum HDL, whereas the amount of apo A-II was not different. We also measured the concentration of apo A-I–containing lipoprotein particles in such medium and observed that there are fewer Lp A-I particles in follicular fluid than in plasma. Lp A-I represented only 36.1% of total apo A-I, compared with 56.5% in plasma (P<.001).
In follicular fluid, HDL-free cholesterol represented only 8.9±0.7% of the total cholesterol, which is significantly different from that measured in plasma HDL (17.4±0.4%, P<.001, not shown). We also calculated the amount of total cholesterol per milligram of apo A-I (Table 2⇓), which was significantly lower in follicular fluid HDL than in serum HDL (750.3±112.3 and 1061.4±29.5 nmol/mg apo A-I, respectively). However, there is only a tendency toward a lower level of esterified cholesterol. Interestingly, in regard to the molecular species of esterified cholesterol, cholesteryl arachidonate was significantly lower in follicular fluid HDL than in plasma HDL and represented 47.7±7.6 and 71.9±7.6 nmol/mg apo A-I, respectively.
The molar ratio of free cholesterol to phospholipid in follicular fluid HDL was significantly lower than that in serum HDL (0.08±0.01 and 0.26±0.02, respectively; P<.001). The fatty acid composition of phospholipid was determined in the phosphatidylcholine fraction, which is predominant among HDL phospholipids in both plasma and follicular fluid. As shown in Table 3⇓, fatty acid species present in follicular fluid HDL phosphatidylcholines are similar to those of plasma HDL, and their amounts, expressed as nanomoles per milligram of total apo A-I, are not significantly different, except for docosapentanoic acid (C22:5, P<.01). Unsaturated fatty acids represented 62.08±0.57% and 58.91±0.72% of total fatty acids in follicular fluid and plasma HDL phosphatidylcholines, respectively.
The relative proportions of lipid components in HDL were slightly different in follicular fluid than in plasma. Compared with plasma HDL, the proportions of free and esterified cholesterol were lower in follicular fluid, whereas the proportion of phospholipids was increased (Table 4⇓).
Kinetics and Dose Responses of Cellular Cholesterol Efflux
We compared the kinetics of cholesterol efflux from the cells with the kinetics from a pool of follicular fluids (n=8) and a pool of homologous sera (n=8) at the following concentrations: 1%, 2.5%, 5%, 10%, 25%, and 50% (vol/vol). The total cholesterol HDL concentrations in each pool represented 0.008, 0.02, 0.04, 0.08, 0.2, and 0.4 mmol/L in follicular fluid and 0.015, 0.037, 0.075, 0.15, 0.375, and 0.75 mmol/L in serum. For each concentration and sample, we measured the fractional cholesterol efflux at different incubation times, which ranged from 10 to 240 minutes. The apparent t1/2 of cholesterol efflux was calculated for each HDL cholesterol concentration from plots of ln(1−FE) versus time, which exhibited linear curves (Fig 1⇓) and were fitted with pseudo–first-order kinetics. Plots obtained from 0.04 mmol/L follicular fluid HDL cholesterol and 0.037 mmol/L serum HDL cholesterol are shown in Fig 1A⇓ and B, respectively. Apparent t1/2 values were plotted as a function of the total HDL cholesterol concentrations used (Fig 2⇓). We show that at equivalent concentrations of HDL cholesterol in follicular fluid and serum, the t1/2 values are in the same range. At cholesterol HDL concentrations above 0.2 mmol/L in follicular fluid and above 0.15 mmol/L in serum, the t1/2 value of cholesterol efflux became independent of acceptor concentrations.
We obtained a series of dose-response curves to determine the relationship between sample HDL cholesterol concentrations and cellular cholesterol efflux. Fig 3⇓ shows results obtained from an incubation time of 4 hours with follicular fluid and serum, with the inset showing a linearization of dose data according to the Lineweaver-Burk plot. As shown in Fig 3⇓, the dose-response curves display the same pattern for the two fluids. The inset on Fig 3⇓ illustrates that the estimated maximal efflux values (Vmax efflux) are not significantly different between follicular fluid and serum (45.9% and 49.6%, respectively), as are the Km values (0.064 and 0.071 mmol/L, respectively). Thus, the efficiency of follicular fluid in promoting cellular cholesterol efflux is similar to that of serum.
Efflux, Esterification, and Transfer of Cell-Derived Cholesterol
On the basis of our previous kinetic and dose-response data, which argue in favor of an essential role of HDL in cholesterol efflux, we compared, at constant HDL cholesterol concentrations (average value, 0.075 mmol/L), the ability of follicular fluids and homologous sera from seven individuals to promote cellular cholesterol efflux. The results are presented in Table 5⇓. The fractional efflux measured from cells to follicular fluids and sera was not significantly different (23.36±1.53% and 24.34±0.60%, respectively). Interestingly, the fractional efflux measured in the presence of VLDL/LDL–depleted serum (as described in “Methods”), at a constant HDL cholesterol concentration, was similar to that of follicular fluids (20.28±0.63% and 23.36±1.53%, respectively). Also, when efflux was expressed as the molar rate of cholesterol efflux per hour and per milligram of apo A-I, there were no significant differences between whole follicular fluid, serum, and VLDL/LDL–depleted serum (18.89±1.07, 23.55±1.78, and 19.05±1.02 nmol/h · mg−1, respectively). In addition, on the basis of constant total cholesterol (0.075 mmol/L, previously used), similar values for cholesterol efflux were observed with isolated HDL from serum and follicular fluid (Fig 4⇓). Similar results were obtained using higher HDL cholesterol concentrations (not shown). These results are consistent with the kinetic and dose-response data and further demonstrate that HDL plays the major role in cellular cholesterol efflux.
The fractional esterification of cell-derived cholesterol, expressed as nanomoles of esterified cholesterol per hour per milligram of apo A-I, was similar in the whole follicular fluid and the VLDL/LDL–depleted serum (3.58±0.31 and 3.23±0.51 nmol/h · mg−1, respectively), whereas it was significantly lower in whole sera compared with follicular fluid (2.32±0.27 nmol/h · mg−1, P<.02). Thus, the presence of apo B/E–containing lipoproteins in the biological medium reduces the rate of esterification of free cholesterol released from cells on the HDL, probably because part of the free cholesterol is exchanged with VLDL and LDL.
In regard to the previous results, we attempted to study the effects of added LDL on the cholesterol efflux and esterification rates mediated by whole follicular fluid. In addition, LDL-supplemented follicular fluids, as esterified cholesterol acceptors, allowed us to estimate the CETP activity in the follicular fluid. Incubation of follicular fluids and LDL-supplemented follicular fluids (used at a concentration of 5%, which corresponds to an average concentration of 0.042 mmol/L cholesterol HDL) were carried out for 4 hours at 37°C (Fig 5⇓). Fractional efflux with LDL-supplemented follicular fluid was significantly higher than that of follicular fluid without LDL (26.06±0.75% and 19.19±0.98%, respectively; P<.001) and reached values obtained with a standard of the sera pool used at 5% (27.04%). This value corresponded to an average concentration of 0.075 mmol/L HDL cholesterol. The cholesterol esterification rate measured in the efflux medium (Fig 5A⇓), expressed as nanomoles of esterified cholesterol per hour per milligram of apo A-I, was significantly lower in the follicular fluid supplemented with LDL (1.94±0.21 nmol/h · mg−1) than in follicular fluid without LDL (4.44±0.4 nmol/h · mg−1, P<.001). In LDL-supplemented follicular fluid, we measured a rate of cholesteryl ester transfer from HDL to LDL of 16.56±1.00%, compared with 23.58% in the pool of sera (Fig 5B⇓). This confirmed the presence of CETP activity in the follicular fluids as well as in sera.
Measurement of Cholesterol Esterification and Esterified Cholesterol Transfer in Follicular Fluid Compared With Homologous Serum in a Cell-Free System
The esterification rate was measured in follicular fluid and in serum devoid of VLDL and LDL by selective precipitation. This procedure was used to compare the esterification rate in HDL only because apo B–containing lipoproteins are absent in follicular fluid. Values of the esterification rate, expressed as nanomoles of esterified cholesterol per hour per milligram of apo A-I, were significantly lower in follicular fluid than in serum (1.15±0.16 and 4.88±0.1 nmol/h · mg−1, respectively; P<.05).
The esterified cholesterol transfer activity, expressed as nanomoles of esterified cholesterol transferred from HDL to added LDL per hour per milligram of apo A-I, was not different between follicular fluid and serum (15.7±5.3 and 12.0±3.2 nmol/h · mg−1, respectively).
The sole presence of HDL in the preovulatory human follicular fluid allowed us to focus on HDL metabolism in an environment devoid of the other lipoproteins. As previously suggested,14 16 17 18 this extravascular fluid might originate, at least in part, from blood plasma transported through the follicle basal membrane, since most of the follicular fluid proteins are plasma proteins. The porosity of the follicle membrane is modulated by hormonal regulation, and in the preovulatory period, the cutoff is about 3×105 d. This size limitation may explain the absence of apo B/E–containing lipoproteins in the follicular fluid, whereas HDL would selectively filter through the follicle membrane. According to Le Goff,16 HDL2 particles are not expected to filter through the follicle membrane. In our fluids, we found that large HDL particles similar to HDL2 represented only 39% of total HDL, compared with 48% in homologous plasma (not shown). Large HDL particles might originate from the remodeling of smaller HDL particles in the follicular fluid, involving proteins as PLTP and hepatic triacylglycerol lipase. Indeed, it has been shown that PLTP has the ability to convert HDL3 particles into two populations of large (HDL2) and small particles (pre-βHDL),25 26 27 28 and recently, mRNAs of PLTP have been detected in large amounts in human ovaries.29 Hepatic triacylglycerol lipase has been detected at the surface of ovaries and also in adrenal glands,30 and its activity was detectable in follicular fluid (X. Collet and B. Jaspard, personal communication). This enzyme may also contribute to the remodeling of HDL in the follicular fluid and in particular to the presence of a relatively high proportion of pre-β HDL particles because we previously showed that pre-β HDL particles represented 18% of total apo A-I in follicular fluid and only 11% in plasma.18 Interestingly, Lp A-I represented only 36.1% of total apo A-I in follicular fluid, compared with 56.5% in plasma (Table 1⇑), and this should be attributed to the lower level of HDL2 particles in follicular fluids rather than the amount of pre-β HDL particles.
LCAT activity measured in a cell-free system was much lower in the follicular fluid than in plasma, and this result is in agreement with those of Le Goff,16 who studied human and equine follicular fluids. In contrast, when fractional esterification was measured on cell-derived cholesterol ([3H]cholesterol released from Fu5AH rat hepatoma cells), the esterification values were significantly higher in follicular fluids than in either homologous sera or in follicular fluid supplemented with LDL. This suggests that in the cell-free system, the amount of free cholesterol in HDL is rate limiting for esterification in the follicular fluid because these lipoproteins are cholesterol-poor particles compared with plasma HDL. In the presence of free cholesterol donor cells, labeled substrate is continuously provided for esterification, and this dynamic system allowed us to show that LCAT activity is not reduced in follicular fluid compared with plasma. Moreover, the supplementation of follicular fluid with plasma LDL led to a marked decrease in the esterification of cell-derived cholesterol. This reduction may be due to the transfer of free cholesterol from the HDL acceptor to LDL, thus removing part of its substrate from the LCAT reaction.
We also demonstrated CETP activity in the follicular fluid. In a cell-free system, the esterified cholesterol transfer, measured as nanomoles of esterified cholesterol transferred from HDL to exogenous LDL, was similar in the follicular fluid and plasma HDL supernatant (isolated by a precipitation procedure).
Taking advantage of the sole presence of HDL in the follicular fluid, we compared the ability of follicular fluids and homologous sera to promote cellular cholesterol efflux using the experimental procedure developed by de La Llera Moya et al.8 For this study we used the Fu5AH rat hepatoma cell line, a well-defined efflux model that exhibited the fastest cholesterol efflux of all cell types studied.2 Our kinetic and dose-response data demonstrated the essential role of HDL in cholesterol efflux since, at similar HDL-cholesterol concentrations, follicular fluid and serum promoted rates of cholesterol release with comparable half-times. For example, the apparent t1/2 value for cholesterol release was 14.8±0.2 hours with serum and 15.9±0.4 hours with follicular fluid (serum HDL cholesterol, 0.037 mmol/L; follicular fluid HDL cholesterol, 0.04 mmol/L). We observed that addition of serum or follicular fluid at higher concentrations did not change the t1/2 value, and this result is in good agreement with those of previous studies,31 which have reported that at infinite acceptor concentrations, t1/2 becomes independent of acceptor concentration and that the rate-limiting step is the desorption of cholesterol from the cell membrane into the unstirred water layer surrounding the cells. In this case, efflux appeared to be a function of plasma membrane composition. Similar conclusions were reached when individual follicular fluids were compared with their homologous sera after 4 hours of incubation with Fu5AH cells.
In addition, the similarity of cholesterol efflux values in follicular fluid and in serum HDL supernatant is in favor of the major role of the HDL in the cholesterol efflux because the removal of apo B/E–containing lipoproteins does not affect the rate of cholesterol efflux. In good agreement with these results, isolated HDL from follicular fluid and serum displayed the same capacity in promoting cholesterol efflux.
According to the model proposed by Fielding and Fielding,32 free cholesterol efflux from the cell membrane to HDL involves at least two types of acceptor: small apo A-I–rich particles, characterized as pre-β HDL particles, which promote fast-released cholesterol from the cell membrane,7 and large α HDL particles, which transport the released cholesterol and are more effective for efflux compared with the small ones on a per-particle basis.4 Although the distribution of HDL particles was somewhat different in the serum and the follicular fluid, the latter being rich in pre-β HDL and relatively poor in large HDL particles,18 the efflux capacities of both media are comparable. The value of fractional efflux measured using Fu5AH cells might reflect the combined effects of both small and large particles and must be regulated by a balance in amount of each type of particle. The selective impact of pre-β HDL particles would be better demonstrated using other cell types or efflux conditions, like short-term incubation on fibroblasts, comparing the efficiency of plasma and follicular fluid to promote the early release of cellular cholesterol.
Interestingly, we also found a significantly higher amount of apo A-IV in follicular fluid than in serum. Apo A-IV acts as a cofactor of LCAT33 but is less efficient than apo A-I and may play a role in the efflux of cholesterol from cells to lipoproteins. Indeed, previous studies34 35 36 have shown that apo A-IV or apo A-IV– and apo AI–containing lipoprotein particles promote cholesterol efflux from various cell lines and have been shown to be associated with LCAT and CETP activities. The apo A-IV was essentially associated with pre-β1 HDL in plasma, whereas follicular fluid apo A-IV was found in VHDL and pre-β1 HDL. A weak signal was also observed in the α HDL3 subfractions (B. Jaspard, unpublished observations). This particle may correspond to the Lp A-I/A-IV or Lp A-IV recently isolated from plasma by immunoaffinity columns. However, the absence of apo-IV mRNA in ovaries37 suggests that apo A-IV lipoprotein particles are not formed in follicular fluid and that they originate through plasma filtration. The specific role of apo A-IV in modulating efflux to follicular fluid is currently under investigation.
At this time, follicular fluid is a unique physiological medium containing only HDL lipoproteins whose use may lead to a better understanding of HDL metabolism.
Selected Abbreviations and Acronyms
|CETP||=||cholesterol ester transfer protein|
|PLTP||=||phospholipid transfer protein|
This work was supported in part by a research grant from ARCOL and Fournier Laboratories, Dijon, France. We acknowledge Professor G.H. Rothblat for valuable discussions and critical review of this manuscript.
Johnson WJ, Bamberger MJ, Latta RA, Rapp PE, Phillips MC, Rothblat GH. The bidirectional flux of cholesterol between cells and lipoproteins: effects of phospholipid depletion of high density lipoprotein. J Biol Chem. 1986;261:5766-5776.
Davidson WS, Rodriguesa WV, Lund-Katz S, Johnson WJ, Rothblat GH, Phillips MC. Effects of acceptor particle size on the efflux of cellular free cholesterol. J Biol Chem. 1995;270:17106-17113.
Mahlberg FH, Rothblat GH. Cellular cholesterol efflux: role of cell membrane kinetic pools and interaction with apolipoproteins AI, AII, and Cs. J Biol Chem. 1992;267:4541-4550.
de la Llera Moya M, Atger V, Paul JL, et al. A cell culture system for screening human serum for ability to promote cholesterol efflux: relations between serum components and efflux, esterification, and transfer. Arterioscler Thromb. 1994;14:1056-1065.
Huang Y, von Eckardstein A, Assmann G. Cell-derived unesterified cholesterol cycles between different HDLs and LDL for its effective esterification in plasma. Arterioscler Thromb. 1993;13(3):445-458.
Reichl D. Lipoproteins of human peripheral lymph. Eur Heart J. 1990;(suppl E):230-236.
Sloop CH, Dory L, Roheim PS. Interstitial fluid lipoproteins. J Lipid Res. 1987;28:225-237.
Caravaglios R, Cilotti RJ. J Endocrinol. 1957;15:273.
Schreiber JR, Weinstein DB. Lipoprotein receptors in steroidogenesis. 1986:359-401.
Vieu C, Jaspard B, Barbaras R, et al. Identification and quantitation of diacylglycerols in HDL and accessibility to lipase. J Lipid Res. 1996;37:1153-1161.
Bligh EG, Dyer WJ. A rapid method of total lipid extraction and purification. Can J Biochem Physiol. 1959;37:911-917.
Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the folin phenol reagent. J Biol Chem. 1951;193:265-275.
Skipski VP, Peterson FF, Barclay M. Quantitative analysis of phospholipids by thin-layer chromatography. Biochem J. 1964;90:374-379.
Tu AY, Nishida HI, Nishida T. High density lipoprotein conversion mediated by human plasma phospholipid transfer protein. J Biol Chem. 1993;268:23098-23105.
Jauhiainen M, Metso J, Palhman R, Blomqvist S, van Tol A, Ehnholm C. Human plasma phospholipid transfer protein causes high density lipoprotein conversion. J Biol Chem. 1993;268:4032-4036.
Pussinen P, Jauhiainen M, Metso J, Tyynelä J, Ehnholm C. Pig plasma phospholipid transfer protein facilitates HDL interconversion. J Lipid Res. 1995;36:975-985.
Lusa S, Jauhiainen M, Somerharju P, Ehnholm C. The mechanism of plasma phospholipid transfer protein-induced enlargement of high-density lipoprotein particles: evidence for particle fusion. Biochem J. 1996;313:275-282.
Hixenbaugh EA, Sullivan TR, Strauss JF III, Laposata EA, Komaromy M, Paavola LG. Hepatic lipase in rat ovary: ovaries cannot synthesize hepatic lipase but accumulate it from the circulation. J Biol Chem. 1989;266:4222-4230.
Rothblat GH, Phillips MC. Mechanism of cholesterol efflux from cells: effects of acceptor structure and concentration. J Biol Chem. 1982;257:4775-4782.
Fielding CJ, Fielding PE. Molecular physiology of reverse cholesterol transport. J Lipid Res. 1995;36:211-228.
Steinmetz A, Utermann G. Activation of lecithin: cholesterol acyl transferase by human apolipoprotein A-IV. J Biol Chem. 1985;260:2258-2264.
Steinmetz A, Barbaras R, Ghalim N, Clavey V, Fruchart JC, Ailhaud G. Human apolipoprotein A-IV binds to apolipoprotein A-I/A-II receptor sites and promotes cholesterol efflux from adipose cells. J Biol Chem. 1990;265:7859-7863.
von Eckardstein A, Huang Y, Wu S, et al. Lipoproteins containing apolipoprotein A-IV but not apolipoprotein A-I take up and esterify cell-derived cholesterol in plasma. 1995.
Duverger N, Ghalim N, Ailhaud G, Steinmetz A, Fruchart J-C, Castro G. Characterization of apo A-IV-containing lipoprotein particles isolated from human plasma and interstitial fluid. Arterioscler Thromb. 1993;13:126-132.